+1 GilPierre, can you please stop cluttering the issues with the copy/pasted yetus artifacts from your local environment?
At least until we decided to use it and have a setup on the official infrastructure.
Thanks, Michael Am 14.02.20 um 16:34 schrieb Gil Portenseigne:
Hello Pierre, It is interesting but i guess that will be the next step. Thanks On Fri, Feb 14, 2020 at 09:04:49AM +0100, Pierre Smits wrote:It seems many projects (including those Apache projects in the Hadoop sphere) use the works of the Apache Yetus (see [1]) project to do pre-commit checks (including IIUC checkstyle, in an automate way) on patch files attached to JIRA tickets and Pull Requests in Github against the latest commit in the branch. This may be helpful to lessen the burden on the contributor. When configured correctly, the test results appear in the ticket. Apache HBASE is such a project using Yetus, See as an example [2] where test results are included through comments of 'Hadoop QA'. Maybe we should consider this for our project? [1] https://yetus.apache.org [2] https://issues.apache.org/jira/browse/HBASE-14498?jql=project%20%3D%20HBASE%20AND%20status%20%3D%20%22Patch%20Available%22%20AND%20resolution%20%3D%20Unresolved%20ORDER%20BY%20priority%20DESC%2C%20updated%20DESC Best regards, Pierre Smits *Proud* *contributor* (but without privileges)* of* Apache OFBiz <https://ofbiz.apache.org/>, since 2008 *Apache Trafodion <https://trafodion.apache.org>, Vice President* *Apache Directory <https://directory.apache.org>, PMC Member* Apache Incubator <https://incubator.apache.org>, committer Apache Steve <https://steve.apache.org>, committer On Thu, Feb 13, 2020 at 9:13 PM Gil Portenseigne < [email protected]> wrote:Hello Michael, Adapting checkstyle configuration is less impacting to our codebase but make us stay different from the java standard. That is the easier path, that will not affect code history. But about getting nearer from the java standard is IMO a nice to have, to make pure java developer feel better discovering OFBiz. I'm kinda prefer the "opposite approach", but we need to discuss if this improvement is worth the history lost. In the example you chose, i see no issue capitalizing module, resource and other. Updating the rule offer the ability to write a constant like : MoDuLe :-). Lots of errors are about line that are 120+ length, missing or uneeded space etc. So that will lead to loads of unrisky modifications. With the usage of IDE with checkstyle plugin this can be done nicely. I check that over the near thousand of java file concerned more than 80% have less than 50 issue... So we can focus our effort onto the big files using a branch to test the idea and ease the issue. And like Daniel said, it is a good subject for a community week effort. I'd really like to be able to efficiently use this feature, in one way or the other, so thank you for starting this discussion. Regards, Gil On Thu, Feb 13, 2020 at 05:44:40PM +0100, Michael Brohl wrote:Hi *, checkstyle currently reports a huge amount of errors. We currently haveanerror count setup in the configuration to prevent the build from failing because of the present errors. Some thoughts/questions to the community: * should we take an approach to fine-tune the configuration so that it better fits the project style? As an example, constants are currently not allowed to be named "module", "resource" etc. which is a common pattern in our code. Changing from ^[A-Z][A-Z0-9]*(_[A-Z0-9]+)*$ to ^[a-zA-Z][a-zA-Z0-9]*(_[a-zA-Z0-9]+)*$ would allow the common naming. The opposite approach would be to rename those to fit the defaultsettings.* should we start an initiative to remove the valid errors like we didwiththe FindBugs initiative some time ago? Thanks for your feedback, Michael Brohl ecomify GmbH - www.ecomify.de
smime.p7s
Description: S/MIME Cryptographic Signature
